368 - Potentially inaccurate weight measurements in an emergency department

Monday, May 1, 2023

9:30 AM – 11:30 AM ET

Poster Number: 368
Publication Number: 368.404

Carrie Daymont, Penn State College of Medicine, Hershey, PA, United States; Hena Choi, Pennsylvania State University College of Medicine, Hershey, PA, United States; Brooke P. Lichak, Pennsylvania State University College of Medicine, Hershey, PA, United States; Eric Schaefer, Pennsylvania State University College of Medicine, Hershey, PA, United States; Kathryn M. McCans, PennState Health Children's Hospital/Hershey Medical Center, Havertown, PA, United States

Presenting Author(s)

Carrie Daymont, MD, MSCE (she/her/hers)

Associate Professor
Penn State College of Medicine
Penn State College of Medicine
Hershey, Pennsylvania, United States

Background: A child's weight is used to guide important decisions in the emergency department (ED), such as dosing high-risk medications and determining the need for admission in infants with poor weight gain. Prior studies of weight errors in the ED have primarily identified extreme errors.

Objective: We sought to estimate how often ED weight measurements are potentially inaccurate by comparing the proportion of ED vs. clinic weights that are >110% of a calculated expected value.

Design/Methods: We performed a single-center retrospective cohort study using electronic health record data for children 0-18 years with ≥1 ED visit and ≥3 clinic visits from 2010-2022. We modified an existing validated algorithm for "cleaning" weight data to determine a child's expected weight at each visit based on their other recorded weights. Recorded weights that deviate from expected may represent a child’s true weight or may be inaccurate. Children seen in the ED may have a lower weight due to illness, but there are few reasons for an acutely higher weight during an ED visit. Therefore, if a larger proportion of ED weights than clinic weights are higher than expected, the majority of these excess high weights are likely inaccurate. The likelihood of inaccuracy is more difficult to determine for clinic weights and ED weights that are lower than expected. Our primary outcome was the difference in the proportion of ED vs. clinic weights >110% of expected. We also evaluated values lower than expected and different degrees of deviation.

Results:

There were 111,453 ED and 499,856 clinic measurements for 42,705 children. The proportion of measurements >110% of expected was 4.13% for ED visits and 2.08% for clinic visits (2.05%, p< 0.001) (Table). For all evaluated degrees of deviation, a larger proportion of ED vs. clinic visits were higher than expected and a larger proportion were lower than expected.

Conclusion(s):

Recorded weights in the ED were more likely than clinic weights to be 5, 10, 20, or 50 percent higher or lower than a calculated expected weight. Some of the deviating values represent a child’s true growth. Others, such as the majority of the 2.1% excess ED weights that are >110% of the expected value, may be inaccurate. Even this relatively low prevalence may be clinically significant. Deviation of +/- 10% can cause harm for children receiving certain high-risk medications. Smaller deviations can impact care for infants with poor weight gain. Further research will address limitations in this analysis and will include clinician assessment of weight accuracy and the clinical consequences of inaccurate recorded weights.